Development of site-specific elastic design spectra for future building codes, Part 1: an updated classification of empirical amplification functions

Abstract

This report illustrates the data collection, processing and elaborations carried out in the framework of strategy 1 (“An updated classification of empirical amplification functions”) of the project “Development of site-specific elastic design spectra for future building codes”. This strategy aims at reconstructing a set of site-specific uniform hazard spectra (UHS), based on seismic hazard inputs representative of the Swiss hazard scenario and comprehensive of local amplification effects. We consider site amplification for a variety of subsurface soil configurations. These site-specific UHSs are sorted according to the characteristics of the corresponding subsoil, and collated with the elastic design spectra from the present building codes. The input empirical data set is constituted by the recordings of the Japanese KiK-net strong motion network, including approximately 700 sites, each instrumented with a downhole reference seismometer (generally encased within the seismic bedrock) coupled with a receiver placed at the soil surface. This instrumental configuration is ideal for the study of site amplification phenomena. Earthquake waveforms from KiK-net stations were processed to derive Fourier amplification functions at each site of the network; these functions are then referred to a common seismic bedrock condition (the same as in the Swiss hazard maps) and finally translated to the response spectral domain, so to express the site amplification in terms of spectral acceleration. In this process, the relevance of non- linear soil behavior is also discussed. At the last processing step, the empirically derived response spectral amplification functions are finally superimposed to seismic hazard inputs characteristic for the Swiss hazard scenario, yielding site-specific UHSs, which are compared with elastic design spectra from the present building code. The obtained results have evidenced the need for an improved site-condition classification and changes in the definition of normative spectra that would take into consideration site-amplification effects, in particular resonances. Conservative and simplified solutions are proposed for the case when only indirect or poor information is available for the considered site; vice versa, in case of well-constrained site characterization, a more specific and complex approach should be preferred. --> This report illustrates the data collection, processing and elaborations carried out in the framework ofstrategy 1 (“An updated classification of empirical amplification functions”) of the project“Development of site-specific elastic design spectra for future building codes”.This strategy aims at reconstructing a set of site-specific uniform hazard spectra (UHS), based on seismichazard inputs representative of the Swiss hazard scenario and comprehensive of local amplificationeffects. We consider site amplification for a variety of subsurface soil configurations. These site-specificUHSs are sorted according to the characteristics of the corresponding subsoil, and collated with theelastic design spectra from the present building codes.The input empirical data set is constituted by the recordings of the Japanese KiK-net strong motionnetwork, including approximately 700 sites, each instrumented with a downhole reference seismometer(generally encased within the seismic bedrock) coupled with a receiver placed at the soil surface. Thisinstrumental configuration is ideal for the study of site amplification phenomena.Earthquake waveforms from KiK-net stations were processed to derive Fourier amplification functionsat each site of the network; these functions are then referred to a common seismic bedrock condition(the same as in the Swiss hazard maps) and finally translated to the response spectral domain, so toexpress the site amplification in terms of spectral acceleration. In this process, the relevance of non-linear soil behavior is also discussed.At the last processing step, the empirically derived response spectral amplification functions are finallysuperimposed to seismic hazard inputs characteristic for the Swiss hazard scenario, yielding site-specificUHSs, which are compared with elastic design spectra from the present building code.The obtained results have evidenced the need for an improved site-condition classification and changesin the definition of normative spectra that would take into consideration site-amplification effects, inparticular resonances. Conservative and simplified solutions are proposed for the case when onlyindirect or poor information is available for the considered site; vice versa, in case of well-constrainedsite characterization, a more specific and complex approach should be preferred.